Last Updated on November 5, 2026 by Team TBH
The question of whether a real-money game app is legal gets answered badly in both directions. One camp treats the whole category as gambling in a thin disguise. The other treats a skill label on a store page as though it settled something. Neither is doing the work, because the actual answer is procedural: there is a test, different states apply it with different thresholds, and the result turns on specific structural features of the product that a player can inspect.
Understanding the test is more useful than memorising a list of states, for the same reason that understanding an exam rubric is more useful than memorising last year’s results. The list changes. The rubric mostly does not.
The predominance test
Federal law does not prohibit skill-based real-money tournaments. Legality is settled at state level, and the dominant framework across US states is the predominance test: does skill or chance predominate in determining the outcome. Not whether chance is present, which it is in almost every competitive activity, but whether it predominates.
That framing has a consequence people often miss. A game containing randomness can still be a skill contest, provided skill is the deciding variable across the outcome. Conversely, a game with no randomness at all can fail the test if the structure around it imports chance in other ways, through mismatched opponents, asymmetric starting positions or results determined by factors outside the player’s control.
As of 2026, roughly 12 US states restrict cash skill gaming, leaving around 38 that permit it. South Carolina prohibits skill games with entry fees outright. Those figures are a snapshot and should be read as one.
What Florida’s four conditions reveal about the rubric
Florida publishes a set of conditions that are unusually explicit and therefore unusually useful as a window into how these assessments work. Four of them: the prize is not composed of the entry fees, the prize does not vary with the fees collected, the operator does not compete for the prize, and the prize is announced in advance.
Read as a group, those conditions are not really about prizes. They are about severing the operator’s interest in the outcome and fixing the terms of the contest before it begins. A prize built from entry fees ties the pot to participation. A prize that floats with fees collected makes the contest unvaluable at the point of payment. An operator that can win has a stake in how the contest runs. A prize announced afterwards leaves the player committing to an unknown.
Each condition closes a route by which a contest stops being a comparison of ability and becomes something else. That is the rubric, expressed as four checkable statements. An operator meeting all four in a state that has not published them is still structurally sound; an operator failing them in a permissive state is relying on the absence of scrutiny rather than on its own design.
Six structural features the test keeps landing on
- Whether the outcome is decided by what the players do. This is the test restated and it is the only question that ultimately matters. Everything else on this list exists because it bears on this one. A format where a player’s decisions and execution move the result is on firm ground; one where they are a minor input is not, whatever it is called.
- Whether both players start from the same position. Equivalent starting conditions, the same board, the same clock, the same objective. Asymmetry that neither player controls is a chance variable introduced by the format, and it pushes the assessment towards the chance side regardless of how skilful the play itself is.
- Whether opponents are matched by ability. Random pairing in a paid contest means a substantial share of outcomes are decided at the moment of pairing rather than during play. Pairing by demonstrated ability is what allows skill to be the operating variable, which is a fairness point and a structural legal point simultaneously.
- Whether the prize is fixed and known in advance. Two of Florida’s four conditions concern this. A prize announced before entry, not composed of the entry fees and not varying with them, means the player is evaluating a defined proposition rather than a variable one.
- Whether the operator stands outside the contest. A service fee taken on entry, with the platform not competing for the prize, places the operator in the position of running the contest rather than participating in it. This is Florida’s third condition and it is the one most directly about incentives.
- Whether identity is verified. Multi-accounting breaks ability-based matching, because one person can hold several ratings or enter the same contest more than once. Verification is therefore part of what makes the skill claim credible, as well as being a financial control. Its absence is more concerning than its presence.
Why the map moved in June
The Pennsylvania Supreme Court placed skill games under the casino slot machine framework in June 2026. That is the single most consequential recent movement in this area and it is a useful lesson in how the ground shifts.
The point is not that Pennsylvania reached a particular conclusion. It is that a state’s position can change through a court ruling, without legislation, and that any list of permissive states compiled before that date carried an assumption that stopped being accurate. An operator’s state availability page should therefore be read as a current statement rather than a settled one, and an operator that acknowledges the map moves is being more accurate than one that presents it as fixed.
How it works in practice
Backspin Games maps onto the structural features above in a way that can be checked against each of them. It publishes ten mobile titles, including 21 Jack, Bingo, Solitaire and Cannon Blast. The entry fee is displayed before the player commits and the prize is stated in advance. Opponents are paired by demonstrated ability rather than at random, and both players face equivalent starting conditions, which addresses the matching and symmetry points together. The platform takes a service fee on entry and does not compete for the prize, which is the operator-outside-the-contest condition. Winnings sit as an account balance until the player completes identity verification, links a payout method and requests a withdrawal. Every title carries a free practice mode running the same rules and interface as the paid version, which means the symmetry and matching claims are observable before any money is involved.
Structural soundness is not a statement about outcomes. No operator can promise a return, and results depend on performance against other players.
A pre-deposit checklist
Five minutes of reading answers most of this. Does the operator publish a list of states it serves, and does it say the list can change. Is the entry fee visible before the commit action. Is the prize stated in advance and described as fixed. Does the operator say anywhere that it takes a service fee and does not compete for the prize. Is identity verification explained, and explained in terms of what it protects.
An operator that answers all five in its own help pages has made the assessment easy, which is itself the finding. An operator that answers none has not been prevented from answering them.
Frequently asked questions
Is a game with any randomness automatically gambling?
No. The predominance test asks whether skill or chance predominates in deciding the outcome, not whether chance is present at all. Many competitive activities contain randomness and remain skill contests because the player’s decisions and execution are the deciding factor across results.
How many states restrict cash skill gaming?
Roughly 12 as of 2026, leaving around 38 that permit it. South Carolina prohibits skill games with entry fees. Those counts are a snapshot: the Pennsylvania Supreme Court moved that state’s treatment of skill games in June 2026, so the figures should be read as current rather than fixed.
Does federal law cover this?
Federal law does not prohibit skill-based real-money tournaments. The determination is made state by state, which is why availability varies and why an operator’s state list is the practical document rather than any national statement.
Why does matchmaking appear in a legal discussion?
Because random pairing in a paid contest means outcomes are substantially decided at pairing rather than during play, which moves weight onto the chance side of the predominance test. skill-based matching with equivalent starting conditions is what lets the result reflect what the players did.
If an app is available in my state, is it definitely lawful?
An operator’s availability list reflects its own compliance assessment, which is informative but not a determination. The more durable signal is structure: fixed prize announced in advance, prize funded separately from fees, operator paid a service fee and not competing, symmetric contests and verified identities.
Learn the rubric, not the list
State lists are useful and they go stale. The predominance test does not, and neither do the structural features it keeps landing on. A player who can look at a format and say whether the prize is fixed, whether the operator competes, whether the starting positions match and whether opponents are paired by ability has effectively run the assessment themselves. That is a more durable skill than knowing which twelve states currently restrict the category, and it survives the next ruling.
To read more content like this, explore The Brand Hopper
Subscribe to our newsletter
